Method for improving operating profits with better automated decision making with artificial intelligence

Abstract

A business method to reverse an authorization request denial made according to general guidelines if the particular customer and the particular transaction pass various threshold tests. Alternatively, each customer is assigned different and independent personal thresholds for transaction types, amounts, times, locations, and context. These thresholds are then applied if an instant payment transaction request is about to be declined.

The invention claimed is: 
     
       1. At least one computer network server for supporting payment authorization services, the at least one server comprising:
 one or more processors; 
 non-transitory computer-readable storage media having computer-executable instructions stored thereon, wherein when executed by the one or more processors the computer-readable instructions cause the one or more processors to—
 receive transaction data corresponding to a plurality of transactions associated with a plurality of transactional entities; 
 compute a fraud score for each of the plurality of transactions at least in part by—
 inputting transaction data of the transaction to a classification model and a smart agent associated with the corresponding transactional entity to generate a first output from the smart agent and a second output from the classification model, the smart agent comprising a representation of historical data of the corresponding transactional entity and the classification model being constructed according to one or more of: data mining logic, a neural network, case-based-reasoning, clustering, fuzzy logic, a genetic algorithm, a decision tree, and business rules, 
 analyzing the first output and the second output to generate the fraud score for the transaction; 
 
 compute a profitability score for each of the plurality of transactional entities; 
 determine that the fraud score corresponding to an increased risk transaction of the plurality of transactions exceeds a first threshold representing an increased likelihood of fraud; 
 determine that the profitability score of a profitable transaction entity of the plurality of transactional entities exceeds a second threshold representing increased profitability; and 
 based on the determinations of increased likelihood of fraud and increased profitability for the profitable transaction entity, issue a transaction approval message. 
 
 
     
     
       2. The at least one computer network server of  claim 1 , wherein generating the first output for each of the plurality of transactions includes comparing contents of at least one data field of the transaction data against a profile of the smart agent associated with the transactional entity corresponding to the transaction. 
     
     
       3. The at least one computer network server of  claim 1 , wherein analyzing the first output and the second output is performed at least in part by retrieving one or more client tuning inputs and incorporating the one or more client tuning inputs into the weighted summation. 
     
     
       4. The at least one computer network server of  claim 1 , wherein, for each of the transactional entities—
 generating the profitability score includes: (a) computing a customer value for the transactional entity, (b) determining an impact on historical transaction levels of the transactional entity, (c) determining a revenue loss assuming cessation of future transactions by the transactional entity, 
 making the profitability determination includes determining that the profitability score entitles the transactional entity to special processing. 
 
     
     
       5. The at least one computer network server of  claim 1 , wherein the computer-readable instructions further instruct the one or more processors to, for each of the plurality of transactions, update the smart agent corresponding to the transactional entity of the transaction according to the transaction data of the transaction. 
     
     
       6. The at least one computer network server of  claim 1 , wherein issuing the transaction approval message for the increased risk transaction is contingent on—
 determining that the profitability score entitles the transactional entity to special processing, 
 determining that a transaction amount of the increased risk transaction is above a third threshold. 
 
     
     
       7. The at least one computer network server of  claim 6 , wherein issuing the transaction approval message for the increased risk transaction is additionally contingent on requiring, based on the transaction amount exceeding the third threshold, the transaction data of the increased risk transaction to satisfy a familiar transaction pattern test. 
     
     
       8. Non-transitory computer-readable storage media having computer-executable instructions for supporting payment authorization services, wherein when executed by at least one processor the computer-readable instructions cause the at least one processor to:
 receive transaction data corresponding to a plurality of transactions associated with a plurality of transactional entities; 
 compute a fraud score for each of the plurality of transactions at least in part by—
 inputting transaction data of the transaction to a classification model and a smart agent associated with the corresponding transactional entity to generate a first output from the smart agent and a second output from the classification model, the smart agent comprising a representation of historical data of the corresponding transactional entity and the classification model being constructed according to one or more of: data mining logic, a neural network, case-based-reasoning, clustering, fuzzy logic, a genetic algorithm, a decision tree, and business rules, 
 analyzing the first output and the second output to generate the fraud score for the transaction; 
 
 compute a profitability score for each of the plurality of transactional entities; 
 determine that the fraud score corresponding to an increased risk transaction of the plurality of transactions exceeds a first threshold representing an increased likelihood of fraud; 
 determine that the profitability score of a profitable transaction entity of the plurality of transactional entities exceeds a second threshold representing increased profitability; and 
 based on the determinations of increased likelihood of fraud and increased profitability for the profitable transaction entity, issue a transaction approval message. 
 
     
     
       9. The non-transitory computer-readable storage media of  claim 8 , wherein generating the first output for each of the plurality of transactions includes comparing contents of at least one data field of the transaction data against a profile of the smart agent associated with the transactional entity corresponding to the transaction. 
     
     
       10. The non-transitory computer-readable storage media of  claim 8 , wherein analyzing the first output and the second output at least in part includes retrieving one or more client tuning inputs and incorporating the one or more client tuning inputs into the weighted summation. 
     
     
       11. The non-transitory computer-readable storage media of  claim 8 , wherein, for each of the transactional entities—
 generating the profitability score includes: (a) computing a customer value for the transactional entity, (b) determining an impact on historical transaction levels of the transactional entity, (c) determining a revenue loss assuming cessation of future transactions by the transactional entity, 
 making the profitability determination includes determining that the profitability score entitles the transactional entity to special processing. 
 
     
     
       12. The non-transitory computer-readable storage media of  claim 8 , wherein the computer-readable instructions further instruct the at least one processor to, for each of the plurality of transactions, update the smart agent corresponding to the transactional entity of the transaction according to the transaction data of the transaction. 
     
     
       13. The non-transitory computer-readable storage media of  claim 8 , wherein issuing the transaction approval message for the increased risk transaction is contingent on—
 determining that the profitability score entitles the transactional entity to special processing, 
 determining that a transaction amount of the increased risk transaction is above a third threshold. 
 
     
     
       14. The non-transitory computer-readable storage media of  claim 13 , wherein issuing the transaction approval message for the increased risk transaction is additionally contingent on requiring, based on the transaction amount exceeding the third threshold, the transaction data of the increased risk transaction to satisfy a familiar transaction pattern test. 
     
     
       15. A computer-implemented method for supporting payment authorization services comprising, via one or more transceivers and/or processors:
 receiving transaction data corresponding to a plurality of transactions associated with a plurality of transactional entities; 
 computing a fraud score for each of the plurality of transactions at least in part by—
 inputting transaction data of the transaction to a classification model and a smart agent associated with the corresponding transactional entity to generate a first output from the smart agent and a second output from the classification model, the smart agent comprising a representation of historical data of the corresponding transactional entity and the classification model being constructed according to one or more of: data mining logic, a neural network, case-based-reasoning, clustering, fuzzy logic, a genetic algorithm, a decision tree, and business rules, 
 analyzing the first output and the second output to generate the fraud score for the transaction; 
 
 computing a profitability score for each of the plurality of transactional entities; 
 determining that the fraud score corresponding to an increased risk transaction of the plurality of transactions exceeds a first threshold representing an increased likelihood of fraud; 
 determining that the profitability score of a profitable transaction entity of the plurality of transactional entities exceeds a second threshold representing increased profitability; and 
 based on the determinations of increased likelihood of fraud and increased profitability for the profitable transaction entity, issuing a transaction approval message. 
 
     
     
       16. The computer-implemented method of  claim 15 , wherein analyzing the first output and the second output at least in part includes retrieving one or more client tuning inputs and incorporating the one or more client tuning inputs into the weighted summation. 
     
     
       17. The computer-implemented method of  claim 15 , wherein, for each of the transactional entities—
 generating the profitability score includes: (a) computing a customer value for the transactional entity, (b) determining an impact on historical transaction levels of the transactional entity, (c) determining a revenue loss assuming cessation of future transactions by the transactional entity, 
 making the profitability determination includes determining that the profitability score entitles the transactional entity to special processing. 
 
     
     
       18. The computer-implemented method of  claim 15 , further comprising, for each of the plurality of transactions, updating the smart agent corresponding to the transactional entity of the transaction according to the transaction data of the transaction. 
     
     
       19. The computer-implemented method of  claim 15 , wherein issuing the transaction approval message for the increased risk transaction is contingent on—
 determining that the profitability score entitles the transactional entity to special processing, 
 determining that a transaction amount of the increased risk transaction is above a third threshold. 
 
     
     
       20. The computer-implemented method of  claim 19 , wherein issuing the transaction approval message for the increased risk transaction is additionally contingent on requiring, based on the transaction amount exceeding the third threshold, the transaction data of the increased risk transaction to satisfy a familiar transaction pattern test.